How we’re rewriting our pricing section for buyers — behind the scenes

Last month a VP Engineering at a Series A company messaged me after visiting clawmetry.com. Their question was blunt: “How much would this cost for my team of 12 running about 35 agents across two environments?”

I answered it in 30 seconds over DM. But our pricing page couldn’t answer it at all.

That’s a problem. The pricing page is where purchase decisions happen. If a VP Engineering has to DM the founder to get a budget number, we’ve lost everyone who didn’t bother to DM.

This is the public gap audit for our pricing section — what it currently communicates, what a buyer actually needs to see, and the five things we’re changing. Same format as the hero section audit we published last month. Same reason: writing the diagnosis publicly forces us to ship the fix, not just discuss it internally.

What a buyer actually sees on our pricing page

Here’s an honest description of how our pricing section reads today from a CTO or VP Engineering’s perspective:

If you’re a solo developer, this reads fine. You install the free tier, you upgrade if you want cloud features. Clean.

But if you’re a VP Engineering evaluating ClawMetry for your platform team, you’re staring at this page asking five questions it doesn’t answer. Here they are.

The five-point gap audit

1 Tiers describe features, not the team outcome

“Multi-node fleet.” “>24h retention.” “Slack/PagerDuty alerts.” These are all accurate feature descriptions. They mean nothing to a VP Engineering trying to answer: will this work for 12 developers running 35 agents across a staging and production environment?

I know the answer is yes. The fleet dashboard handles exactly this. But nowhere on the pricing page does it say: Pro gives your whole platform team a single dashboard across every agent, every environment, with per-session cost attribution by developer and by model.

Buyers buy outcomes. Feature lists are for the evaluation phase, after the buyer has already decided to look deeper. The pricing page is earlier than that — it’s the should I even evaluate this? moment. Lead with what they get, not with what the product does.

2 The data residency moat is buried

This is the one I lose the most sleep over. ClawMetry’s strongest enterprise differentiator is the architecture: local-first compute on DuckDB, E2E encrypted cloud sync, your agent conversations stay under your key and never reach our servers in readable form. This is why regulated-industry teams (FinTech, healthcare, defense contractors) choose us over LangSmith or Datadog.

Those buyers have a procurement checklist. One item is always: “Where does our data go?” If the answer is “a third-party SaaS” — deal dead. Our answer is: “It doesn’t leave your infrastructure in a form we can read.” That’s the moat.

But right now that answer lives in a small badge (“E2E Encrypted”) above the tier comparison. A compliance officer scanning a pricing page won’t catch it. It needs its own callout — ideally with a link to the data processing architecture — so the procurement checklist can be checked without a call with me.

3 No TCO anchor — the cost of not monitoring

No CTO approves a budget line for an observability tool based on the tool’s price. They approve it based on ROI: what does NOT having this cost us?

Here’s a concrete number I can use: one overnight runaway loop — an agent stuck in a retry cycle from midnight to 6 AM — burns roughly 1.5–3M tokens depending on the model. At current rates, that’s $300–$600 of wasted spend in a single incident. ClawMetry detects this in real time and can trigger a Slack alert within seconds of the loop pattern appearing. If you run agents at scale, one caught loop pays for a year of Pro.

That math should be on the pricing page. Not as a “we’ll save you money” marketing claim, but as a concrete ROI anchor a VP Engineering can drop into a budget conversation: we caught two runaway loops last quarter, that’s $800 recovered, the tool costs less than that per month.

Right now the pricing page has no frame for this. You see a price. You don’t see what the alternative costs.

4 No path for procurement

A 200-person company can’t expense SaaS tools on a personal credit card. Procurement needs: an invoice or PO path, a data processing agreement (DPA), and answers to the security questionnaire before anyone signs off.

Our pricing page has a “Get started” button. It does not have:

• An option to pay via invoice/PO for annual plans
• A DPA or security posture link (SECURITY.md is in the GitHub repo — most buyers won’t find it)
• A contact path for legal or procurement questions that isn’t the generic support email

I’ve lost at least two enterprise evaluations not because ClawMetry was the wrong tool, but because procurement couldn’t find answers to questions they needed answered before they could approve the spend. That’s fixable with copy and a few links — not a product change.

5 Scale ambiguity kills the budget conversation

The VP Engineering who messaged me had a specific question: “How much for 12 developers running 35 agents?” The ambiguity in our pricing is whether the unit is per developer, per agent, or per node. These give wildly different numbers. If it’s per agent, 35 agents is expensive. If it’s per node (per deployment machine), their setup might be two nodes — and it’s very cheap.

The actual answer: Pro is per workspace, not per developer or per agent. One OpenClaw workspace running on one machine = one node. Most teams with 12 developers share 2–3 nodes (staging, prod, dev). That’s a very different budget conversation than “$X per developer per month.”

The pricing page needs to state this plainly. Not in a tooltip. Not in an FAQ five scrolls down. In the pricing tier description, the first time someone reads it.

Before/after: feature-list tiers vs. outcome-framed tiers with scale clarity. The tier structure doesn’t change; the framing does.

What the rewrite will say

This is the working draft. Not finalized — same as the hero audit, I’m publishing the draft so there’s a diff to hold us accountable.

The tier structure doesn’t change — free OSS, paid cloud. But every line in the Pro tier now answers a buyer question rather than listing a feature. “Fleet view” becomes “One dashboard for your platform team.” “Alerts” becomes “catches the overnight incident before it costs you $600.” The per-node pricing is stated plainly in the tier description, not buried.

The data residency callout moves from a badge to a first-class section. The PO/DPA line is one sentence, but it tells procurement: this vendor has thought about how enterprise companies buy software.

What stays the same

The free OSS tier stays exactly as-is. 120K+ installs came through the OSS funnel, and the frictionless pip install clawmetry install is the right entry point for individual developers. We don’t change that funnel at all.

The features themselves don’t change either. This audit is purely about framing. The alert system, the fleet dashboard, the E2E encryption — all of it exists and works. The only thing broken is how we describe it to someone with budget authority who landed on the page from a Google search for “AI agent observability enterprise.”

Why we’re publishing this before shipping it

Two reasons, same as last time.

The first is accountability. Writing “we’re rewriting the pricing page” publicly means we actually do it. The copy draft is above. The only step left is a PR. That PR ships this week.

The second is that this pattern — developer-tool-meets-enterprise-buyer, pricing written for the builder not the procurement team — is universal. If you’re a founder in the same spot, maybe the five gaps above are a useful checklist for your own pricing page audit. The questions a VP Engineering can’t answer from your pricing page are usually the same questions: what’s the scale unit, where does data go, how do we buy this with a PO, and what does it cost us to not have this?

We built ClawMetry because I needed it running OpenClaw agents around the clock. The OSS install is how we grew to 120K+ installs across 100+ countries. But the cloud product — fleet dashboards, multi-node visibility, Slack/PagerDuty alerts, approval workflows — exists for platform teams making a deliberate purchasing decision. Those buyers deserve a pricing page that talks to them in their language. This rewrite is that page.